Faster R-CNN training problems encountered in their own data sets

Disclaimer: This article is a blogger original article, follow the CC 4.0 BY-SA copyright agreement, reproduced, please attach the original source link and this statement.
This link: https://blog.csdn.net/lianggyu/article/details/100081350

Code links: https://github.com/endernewton/tf-faster-rcnn

(1)TypeError: write() argument must be str, not bytes:

          lib / datasets / voc_eval.py line 105 to

cachefile = os.path.join(cachedir, '%s_annots.pkl' % imagesetfile.split("/")[-1].split(".")[0])

          Line 121 is modified

with open(cachefile, 'wb') as f:

(2)InvalidArgumentError (see above for traceback): Assign requires shapes of both tensors to match. lhs shape= [4096,21] rhs shape= [4096,3]

         tools / demo.py line 141 to their own data sets 21 (the number of classes 1), such as the category number of 10, on to

net.create_architecture("TEST", 11, tag='default', anchor_scales=[8, 16, 32])

(3)This application failed to start because it could not find or load the Qt platform plugin "xcb"

         Installation pyqt library: conda install pyqt

(4)ImportError: No module named pycocotools.coco

        If the data set made into voc format, pycocotools commented on in the code. In the 23 to 25 lines of /lib/datasets/coco.py:

#from pycocotools.coco import COCO
#from pycocotools.cocoeval import COCOeval
#from pycocotools import mask as COCOmask

(5)TypeError: load() got an unexpected keyword argument 'encoding'

         Delete the data / cache before training; testing before deleting data / VOCdevkit2007 / annotations_cache

 

Guess you like

Origin blog.csdn.net/lianggyu/article/details/100081350